By Makar Evdokimov - Data Integrity Senior Associate
At Vencon Research, we pride ourselves on being the leading specialist provider of compensation and pay metrics to the global professional services, management, and IT consulting industries.
In this article we take a closer look at the statistical measures used in our compensation reports to represent remuneration levels in the market. Taken together these ensure accuracy and relevance for a variety of benchmarking purposes.
Mean: a common starting point with considerable limitations
Looking at the average (mean) salary might be the most intuitive way to get an idea about the overall situation in the market. Statistically speaking, the arithmetic mean is an unbiased estimator of expected value, so from a theoretical perspective, it is a meaningful way to describe the distribution of values with a single number. However, in practice, the average may be somewhat misleading, especially when dealing with remuneration.
Even if we compare employees that have very similar scope of responsibilities and hold positions in firms which are direct competitors, the financial compensation may vary drastically from one case to another. Quite often, such a population will include a few individuals whose remuneration is substantially higher than that of the rest of the group. Such incumbents would be the outliers that drive the average value for the population upwards. The average considers all remuneration values in the population, but it may still be considerably higher than the financial compensation of most of the employees. This undermines the ability of the mean to represent the level of remuneration in the market. Therefore, it is common practice to use other, more robust indicators to aggregate remuneration data in a meaningful way.
Median and percentiles: analysing data distribution
The median is the most common measure used when evaluating remuneration levels in a market. While the mean requires an arithmetic operation involving all values in the sample, the median is rather defined by how the values in the sample are distributed across its range. The median value of remuneration in a population is basically such a value that half of the population is paid lower than the median and the other half is paid higher.
The median provides meaningful insight into the distribution of the values and it is more robust to outliers than the mean. This makes it one of the most important measures utilized in our surveys. Nonetheless, a single value can only say so much about a large dataset, so to provide an extensive overview of the market distribution of various remuneration components, the entire percentile range alongside the median (50th percentile) is presented in our surveys. This is especially helpful if your firm is interested in paying employees above or below the market, which means that the target market percentile (TMP) is not the median but a different point within the range. Vencon Research surveys provide all the data and tools necessary for benchmarking against any target - allowing you to select market target percentiles ranging from the lowest 5th to the highest 95th.
Our surveys also display distribution in the form of boxplot visualizations, charts that focus on five data points in the population: minimum, 25th percentile, median, 75th percentile, maximum.
Such visual aid gives a quick understanding of how the distribution is shaped which can become a valuable insight about the market.
Midpoint: a range-focused indicator
The midpoint (also known as the mid-range) is an arithmetic mean of the maximum and the minimum values in the population.
Even though it seems to be a poor measure of the remuneration level in the market, since it is very much not robust to outliers, the midpoint value is a range-focused indicator that can tell you how much the values of remuneration are dispersed in the market.
Compa-ratio: a key benchmark for salary comparison
One of the most useful metrics utilized in compensation benchmarking is the compa-ratio. The compa-ratio is calculated as your firm’s pay level divided by the market pay level. The selected measure of pay level can vary: mean, midpoint, median, or any other percentile can be used. Nonetheless, it is most common to calculate compa-ratio based on medians. In this case, a compa-ratio value of 0.85 or 85% indicates that the level of remuneration at your firm is 15% below the market level, while a compa-ratio of 1.1 would mean that your employees are paid 10% more than in the market. The comparative ratio assesses how competitive the remuneration at your firm is and gives a specific quantitative estimate of how far your current pay level is from your target in relative terms.
Our reports include the compa-ratio in interactive tool format, allowing you to select the target market percentile with which to compare your firm’s compensation. This allows you to quickly interpret the difference between your firm and the market pay rate at any level or sub-level, across while targeting any point on the entire percentile range.
Actionable statistics for accurate and goal-oriented benchmarking
In our comprehensive compensation benchmarking surveys, Vencon Research utilizes these statistical measures to deliver precise and insightful analyses. By comparing your firm’s pay levels with market data, you can determine how competitive your compensation packages are. At Vencon Research, we are dedicated to providing our clients with the tools and insights needed to make informed decisions about employee compensation. Our meticulous approach ensures that you have the most accurate and actionable data at your fingertips, helping you maintain a competitive edge in the market.
To find out more about our surveys and our benchmarking methodology do not hesitate to get in touch. Our team is always ready to provide personalized assistance to meet your specific needs.